The Dynamics of Abyssal T-phases

Ralph Stephen, Geology & Geophysics

DOEI Project Funded: 2001

Proposed Research

Earthquakes at ridges and transform faults under the deep ocean excite energy in the SOFAR (SOund Fixing And Ranging) channel that can be readily observed on hydrophones suspended in the ocean and on seismometers near the coast. Arrivals that travel this path are called T-phases, or tertiary phases, because they are the third principal arrival on the time series recorded at a station. (The primary (P, first) and secondary (S, second) arrivals that are commonly observed from oceanic and continental earthquakes travel as seismic body waves through the solid earth all the way from the source to the receiver.) "Abyssal T-phases" are generated by earthquakes in the deep (>3000m) ocean and "slope T-phases" are generated at continental margins. Because of the efficiency of T-phase propagation in the SOFAR channel, hydrophones in the deep ocean can sometimes detect much smaller earthquakes than a seismic sensor at comparable range. The arrival time of abyssal T-phases has been used extensively to locate small earthquakes on the Juan de Fuca Ridge, the East Pacific Rise and the Mid-Atlantic Ridge. At least one swarm of abyssal T-phase identified earthquakes has been associated with volcanic and hydrothermal activity and there is tremendous potential to apply abyssal T-phase data to study geological processes in the deep sea. A major impediment to the application of abyssal T-phases, however, is that the physical mechanism(s) for getting energy from the earthquake into the SOFAR channel are unknown. Although arrival time (kinematic) information can be used to approximately locate an event, an understanding of the dynamics of T-phase generation will be necessary to provide accurate locations and to infer earthquake source mechanisms, magnitudes and depth. This will lead to improved interpretations of geological phenomena. Since earthquakes in oceanic crust have different characteristics than earthquakes in continental crust, comparing the two can also lead to improved models for earthquake generation.

Final Report

This is the final report for my DOEI Research Grant, "The Dynamics of Abyssal T-Phases" which ended June 30, 2003. The characteristics of earthquakes, as revealed by T-phase observations, will provide important constraints on physical models of crustal processes under the oceans. Since T-phases propagate to long ranges in the ocean relatively few observing stations (about six) at large separations (about 2000km) are necessary to identify and approximately locate earthquakes in a whole ocean basin (like the North Atlantic). We do not know however, how many stations or how closely spaced they need to be in order to infer earthquake source mechanisms, magnitudes and depth because we do not know the physical mechanisms responsible for getting T-phase energy into and out of the SOFAR channel. The goal of this project was to quantitatively determine and parameterize the physical mechanisms responsible for T-phase generation. We appreciate the support from the DOEI that provided "seed money" to get us started in the exciting field of earthquake acoustics (T-phases).

Earthquakes at ridges and transform faults under the deep ocean excite energy in the SOFAR (SOund Fixing And Ranging) channel that can be readily observed on hydrophones suspended in the ocean and on seismometers near the coast. Arrivals that travel this path are called T-phases, or tertiary phases, because they are the third principal arrival on the time series recorded at a station. (The primary (P, first ) and secondary (S, second) arrivals that are commonly observed from oceanic and continental earthquakes travel as seismic body waves through the solid earth all the way from the source to the receiver.) "Abyssal T-phases" are generated by earthquakes in the deep (>3000m) ocean and "slope T-phases" are generated at continental margins. Because of the efficiency of T-phase propagation in the SOFAR channel, hydrophones in the deep ocean can sometimes detect much smaller earthquakes than a seismic sensor at comparable range. The arrival time of abyssal T-phases has been used extensively to locate small earthquakes on the Juan de Fuca Ridge, the East Pacific Rise and the Mid-Atlantic Ridge. At least one swarm of abyssal T-phase identified earthquakes has been associated with volcanic and hydrothermal activity and there is tremendous potential to apply abyssal T-phase data to study geological processes in the deep sea. A major impediment to the application of abyssal T-phases, however, is that the physical mechanism(s) for getting energy from the earthquake into the SOFAR channel are unknown. Although arrival time (kinematic) information can be used to approximately locate an event, an understanding of the dynamics of T-phase generation will be necessary to provide accurate locations and to infer earthquake source mechanisms, magnitudes and depth. This will lead to improved interpretations of geological 1/3 phenomena. Since earthquakes in oceanic crust have different characteristics than earthquakes in continental crust, comparing the two can also lead to improved models for earthquake generation.

Under funding from this grant I worked with Debbie Smith and Clare Williams, an MITWHOI Joint Program student, to quantitatively study T-phase events from the mid-Atlantic Ridge at the Kane and Atlantis Transform Faults from 1999 to 2001. Debbie had acquired raw acoustic data from the six element hydrophone array deployed in the North Atlantic by the NOAA-PMEL group. In order to study this hydrophone data in the same fashion as our broadband seismic data it was necessary to write MATLAB code: a) to scan the whole data set for T-phase arrivals (and other noise sources such as whales and airguns), b) to display time series and spectra for single noise events, c) to display the T-phase arrivals (as time series and spectra) as they appear on all six hydrophones, d) to display the data in a format similar to the NOAA-PMEL format for comparison with their results, and e) to write a ray tracing code to predict bathymetric blockage in the presence of rough topography. All of these MATLAB codes were written by me under the support of the DOEI.

This DOEI supported project laid the ground work for four initiatives:

1. Clare Williams used the Matlab code as an important tool in her analysis of T-phase events at the Atlantis and Kane Fracture Zones on the Mid-Atlantic Ridge. Clare presented her analysis at the 2003 Spring AGU (American Geophysical Union) meeting in Nice (Williams et al, 2003) and at her Generals Examination in September 2003 (Williams, 2003). She will also be presenting a paper at the Spring 2004 ASA (Acoustical Society of America) meeting (Williams et al, 2004). In addition she has prepared a technical report outlining her analysis in detail which can be used by subsequent students. The questions addressed in her study were:
   
   
   • Does water depth at the event location influence the characteristics of the T-phase?
     
   • Does the range (ie distance from the event to the hydrophone) of the T-phase propagation path influence the characteristics of the T-phase
     
   • Does bathymetric blocking of the acoustic energy occur along the propagation path due to the bathymetry of the Mid-Atlantic Ridge?
     
   • Is it possible to define the T-phases in our study areas as “slope” or “abyssal” T-phases?
   
   
2. Debbie Smith and I prepared three proposals to the National Science Foundation for further funding to study T-phase dynamics (Stephen and Smith, 2001, 2002a, 2002b) . Although none of these were funded they did raise some important questions regarding the physics of T-phases to the community.
   
   
3. Bob Odom, from the University of Washington, and I have received funding from the National Science Foundation and the Office of Naval Research to hold a workshop at WHOI in March 2004 on "Seismo-acoustic applications in marine geology and geophysics". This workshop is also being partially supported and sponsored by the DOEI. More information on the workshop is available at: http://www.whoi.edu/institutes/doei/activities/activities.htm
   
   
4. Bob Odom and I are also co-chairing a Special Session at the Spring ASA (Acoustical Society of America) meeting in New York in May 2004 on the "Ocean Acoustics of Earthquakes".

Stephen, R.A. and Smith D.K., The dynamics of abyssal T-phases, NSF proposal, August 2001.

Stephen, R.A. and Smith D.K., The dynamics of abyssal T-phases, NSF proposal, February,2002a.

Stephen, R.A. and Smith D.K., Topographic blocking of T-phases in the North Atlantic, NSF proposal, August 2002b.

Williams, C.M., Stephen, R.A. and Smith, D.K. Are T-phases blocked by bathymetry? A study of seismic events located at the Kane and Atlantis transform faults. Spring AGU, Nice, France, 2003.

Williams,C.M. Hydroacoustic events located near the Atlantis (30ºN) and Kane (23º30’N) Transform Faults on the MAR. Generals Exam Paper and Powerpoint Presentation, WHOI/MIT Joint Program, 2003.

Williams, C.M., Stephen, R.A. and Smith, D.K. Williams,C.M. Hydroacoustic events located near the Atlantis (30ºN) and Kane (23º30’N) Transform Faults on the MAR. Spring ASA, New York, NY, 2004.

Originally published: January 1, 2001